1. Field of the Invention
The invention relates to antennas for use with portable and other computing devices, such as laptop computers. More specifically, it relates to antennas that may be part of removable components such as PCMCIA (personal computer memory card international association) cards or the like that provide wireless communication to the computing devices.
2. Description of the Related Art
Some computing devices, such as laptop computers, may not be manufactured with wireless communication capability. Rather, some of these devices may have slots or similar coupling locations into which wireless communication devices may be mated to provide the host computing device with wireless capability. The wireless communication device can be for example a PCMCIA (personal computer memory card international association) card, and can include a transceiver and other circuitry coupled to an antenna and matable with the host device to provide wireless communication capability thereto. While explained herein in terms of a laptop computer as the host device, and a PCMCIA card as the wireless communication device, it will be appreciated that the invention is not so limited, and other host devices, such as PDAs and desktop computers, and other wireless communication devices for establishing wireless communication through a cellular network or through Bluetooth, WiFi and other types of wireless links and channels are also contemplated.
Diversity antennas used with wireless communication devices, especially portable and mobile devices, are very beneficial in improving the quality of the received signal in a wireless communications receiver. Typical diversity antenna systems consist of a main antenna and a diversity antenna, although there could be more than one diversity antenna. The initial benefit of diversity comes from the de-correlation of the fading between two separate antenna systems. The antennas can be spatially separated and/or use orthogonal polarizations (i.e. vertical and horizontal polarizations, right and left circular polarization, etc.) During a fade, the signal strength is degraded to the point that long error bursts occur in the received signal, severely degrading the overall received radio throughput, amongst other degradations. Diversity helps alleviate this problem by having two antennas separated in space and/or polarization, providing two nearly independent receive signal channels or paths which do not experience fades in the same way (that is, they are de-correlated). Thus while one antenna may experience a deep fade the other antenna may be within 3 dB of its nominal signal level. The result of this is that links with rapid fading that can go −15 dB or more below the average signal strength in a fade on a single channel system (non-diversity) but may be reduced to only −4 dB or −5 dB below the average signal strength with diversity on a statistical basis. In this example, diversity would provide an effective gain of 11 dB to 10 dB. Thus the reduced loss of signal prevents the channel from being dropped far less frequently than it would with a single deep fading channel. Typically the diversity antenna may be separated by as little as one eighth of a wavelength and still experience a significant gain over a single channel non-diversity antenna.
FIG. 1a shows a simple two antenna diversity system used on a PCMCIA (personal computer memory card international association) card 10 in a laptop computer 12, in which two vertical dipoles or monopoles (11 and 13) are employed. In FIG. 1b, an orthogonal dipole/monopole configuration is shown that uses a vertical monopole/dipole 14 with horizontal monopole 15 disposed normal to the side of the laptop case. In FIG. 1c, the horizontal monopole is replaced with a PIFA (Planar Inverted-F Antenna) style antenna 17 and a vertical antenna 16. Both the PIFA and the horizontal monopole use the laptop case as the “counterpoise” for the associated antenna system. An “antenna counterpoise” is a virtual ground for balancing the currents in the antenna by establishing a zero reference potential for feeding the active antenna element. It can be any structure closely associated with (or act as) the ground which is connected to the terminal of the signal receiver or source opposing the active antenna terminal, (that is, the signal receiver or source is interposed between the active antenna and this structure). The “antenna counterpoise” may be directly, capacitively or inductively coupled to the surrounding ground plane if there happens to be one there.
One of the main disadvantages of these sample diversity systems is the generally poor isolation between the antennas, sometimes as low as a few dB but typically only 6 dB. With diversity isolations greater that 10 dB being preferred, consideration may be given to improved orthogonality between these antennas to increase the diversity isolation. Higher diversity isolation essentially means less correlation between the separate antennas and therefore a reduced probability of destructive interference or fading.
Another consideration is the interactions between a dipole like-antenna and an orthogonal dipole/monopole with a substantially symmetrical geometry normal to the main dipole length vector. Small form factor wireless communications devices, such as PCMCIA cards, provide very limited external space to include antennas with high efficiency, wide bandwidth, multiple bands and diversity all at the same time. This tight space constraint results in interaction between the various antenna elements, even if the antennas have good isolation between the selected paths or “ports.” This is further complicated by the interaction between the various antenna systems and the computer or platform to which the card is mated.
Thus one consideration is the fabrication of a high performance main and diversity antenna system for use in a PCMCIA card, with the aim of achieving good antenna efficiency with high isolation between the main and diversity antennas and high isolation between the main antenna and the radiated self-noise from the host device (for example lap top computer), while maintaining an acceptable industrial design (ID) appearance. These results are ultimately reflected in the Total Isotropic Sensitivity (TIS) and Total Radiated Power (TRP) performance of the antenna.
Optimum dipole location for minimum laptop self-noise is another consideration. Laptop computers have traditionally been designed primarily for user computer functionality and conformity with FCC part 15 regulations. In more recent times, functionality has been expanded to include wireless network connections such as cellular communications and WiFi. Since the FCC part 15 requires only radiated noise limitations, the issue of self-noise for added or integrated wireless network solutions has not been considered. Consequently, while compliance with FCC part 15 has been achieved, there are high levels of RF surface currents and RF voltage antinodes all over laptop computers. Furthermore, laptop computers now can have prescribed locations at which PCMCIA cards and similar devices can be added after-market, and these locations have become the location for accommodating wireless solutions. The concern is that self-RF noise generated or reaching in these locations de-senses the receiver part of the transceiver. Radiation in the PCMCIA slot regions may be substantially vertically polarized, and conduction currents from the laptop chassis generate conduction noise into antenna structures, such as the traditional monopole, that use the chassis as the substantial counterpoise for the antenna. This latter case can be the main mode of self-noise for PCMCIA-based wireless modems. The lowest noise is generated in the region of the PCMCIA slot in the E-field direction parallel to the long edge of the slot opening in the laptop.
FIGS. 12a-12c show a typical laptop computer with a PCMCIA or other PC card slot 1201 in the side wall of the laptop 1200. The electric fields Ex, Ey and Ez are shown as indicated. FIGS. 12b and 12c show conventional antenna configurations as currently employed. The extension of the PCMCIA card 1206, 1206′ outside of the slot is shown. In FIG. 2b, the antenna 1207 is in the form of a typical monopole antenna that uses the laptop chassis as its counterpoise, and the antenna 1208 is in the form of a vertical antenna that can be a monopole that uses the chassis as its counterpoise, or the antenna could be made longer and configured as an end-fed dipole that is only weakly coupled into the chassis. Antenna 1209 is in the form of a PIFA (Planar Inverted-F Antenna). This style of antenna excites currents and voltage antinodes in the associated ground plane, which acts as a counterpoise to the PIFA. Put simply, the PIFA is a wide monopole that excites the ground plane.
With the exception of the end-fed dipole antenna, all of the antennas of FIGS. 12a-12c suffer from conducted RF noise from the chassis or ground plane of the computer. The end-fed dipole operates best when excited at a low impedance point on the ground plane or chassis. Since this dipole antenna is end-fed, it represents a very high impedance to the ground plane and hence reduces the conducted noise to the dipole.
Optimum dipole location and shape for maximum bandwidth in a small volume is another consideration. Almost all laptop computers today have at least one slot available for mating a PCMCIA card or similar device to the laptop computer. The extent of the projection of a PCMCIA card outside the slot in the side of the laptop is primarily limited by aggressively small industrial design (ID) constraints that have little concern for the needs of RF antenna functionality. Additional constraints are imposed by the mechanical enclosure and its requirements for welding line wall thickness and studs and so forth.
The size of an antenna enclosure has the greatest influence on the antenna performance at the lowest required operating frequency. For an ideal fat dipole the optimum length is 0.45λ, with λ being the wavelength of the interest. However, for cell-phone applications, adequate performance can be achieved with top-loaded dipoles or fat dipoles with a length as short as 0.30λ. Antennas as short as 0.125λ require significant top-loading and often require sophisticated matching circuits to achieve the necessary bandwidth.
In addition, the location of a dipole antenna near a significant ground plane also impacts the bandwidth and performance of a dipole antenna. By way of example, a Yagi antenna requires a minimum separation of reflector from the driven element (typically a dipole) of 0.04λ. The optimum separation is 0.15λ to 0.25λ with adequate performance as close as 0.09λ. As the separation decreases below 0.25λ, the front to back ratio decreases to unity and the bandwidth also decreases.
By way of example, Novatel™, in the C110 Type II PCMCIA card, uses a Yagi style antenna with the ground plane of the PCMCIA card as the reflector, a balun-fed dipole, and a director in order to operate above 1.90 GHz in a cellular application. The spacing between elements is nominally at the minimum of 0.04λ as a result of needing to fit within an overall length of 22 mm. This antenna is integral with the main PCB (printed circuit board) and requires no external antenna components. The folded nature of the antenna elements reflect the struggle to achieve a match even at this high frequency, let alone attempting a solution at 0.824 GHz. The very nature of this three-element Yagi design renders a 0.824 GHz solution extremely inefficient and/or limited bandwidth.
Optimum dipole location and style for minimum specific absorption rate (SAR) in a small volume is yet another consideration. SAR is a direct measure of the amount of RF power absorbed into human tissue due to a transmitting device in close proximity to it. This is a particularly important mobile phone issue as the transceivers of the device are employed in close proximity to the operator's head. The required standards and conditions for the measurement of SAR are defined and regulated by the FCC. There are several basic approaches to SAR reduction:                1. Reduce the radiated RF power        2. Place a screen between the radiator and the tissue        3. Place a resonant reflector between the radiator and the tissue        4. Use an antenna design with a significant front-to-back ratio, pointing the null towards the tissue.        5. Increase the separation distance between the radiator and the tissue        6. Spread out the surface current more over the radiator, particularly close to the radiator feed point in the case of a dipole or a monopole.        
While these seem like simple remedies they each come with a cost, and a trade-off is required that usually impacts either industrial design (ID) and/or antenna and system performance.
Most traditional PCMCIA or PC cards are designed with a single PCB in mind, with antenna assemblies added to the outside edge of the card. The antenna elements typically comprise monopole antennas, whip antennas or PIFA (planar inverted-F antenna) antennas. Some use a coplanar dipole as the radiator, but this has been the choice of expedience of parts and of having a minimum vertical profile. This latter application, if used at all, has mostly been used at 1.8 GHz and above, due the unacceptable size of the antenna at lower frequencies such as 850 MHz.
The SAR “hot spot” most typically occurs close to, if not directly under, the feed point for the antenna. FIG. 15a shows a normal monopole 1501 used in a PCMCIA card 1520 in a laptop computer 1500. The antenna feed point 1502 is at the intersection point between the laptop 1500 case and the monopole 1501. Directly below this point and in the near surface of the tissue of the operator is where the “hot spot” 1503 will usually be found. The same result occurs for a whip antenna, whether normal or vertical at the feed point. The use of a standard dipole 1504 is shown in FIG. 15b, with its associated SAR “hot spot” at 1505 beneath the dipole. In some cases a Yagi antenna can be used in lieu of the dipole, to achieve some SAR reduction.
Inductive coupling between the antenna assembly and the printed circuit board (PCB) is another consideration. In some situations, it may be desirable to use such inductive coupling. An inductive coupling arrangement can be useful with air core transformers having only a few turns on both primary and secondary sides, for instance. However, such air-cored transformers have significantly more flux leakage than a high Mu ferrite-cored transformer. This flux leakage constitutes the uncoupled magnetic flux that does not pass through both coils. The consequence of this leakage is to produce an uncoupled inductance called leakage inductance. This acts in series with both the primary and secondary sides of the transformer, whereas the common inductance is called the mutual inductance and accounts for the magnetic field that is accepted by both sides of the transformer. While the leakage inductance is often perceived as loss, it is in fact conservative and can be cancelled out by using series capacitance or shunt capacitance. The main issue is that if the leakage (uncoupled) inductance exceeds the mutual inductance, the capacitive tuning required will result in a narrower band coupling.
The simplest design rule to minimize the flux leakage is to widen the trace width and to push the two windings as close together as possible. Once the gap-to-width ratio drops below 0.2, the leakage inductance becomes much less that the mutual inductance.
An advantage of the use of inductive coupling is that it simplifies the interconnection between two RF circuits, which, in the case of an antenna assembly, is between the PCB containing the bulk of the circuitry and the FPCB (flexible printed circuit board) of the antenna element(s). The inductive coupling eliminates the need for direct soldering, coaxial connection, zif sockets or pogo pins, etc. The perceived disadvantage is the leakage inductance and the size of the coupling loops, which is directly related to the maximum operating wavelength.
Reference is first made to FIG. 16a, in which conventional arrangement in which a balun 1601 on a main PCB 1630 is used to drive the antenna (not shown) or other balanced device in a differential manner. Balun 1601 is connected directly to a balanced antenna feed system 1603 via a gap port 1602. This interconnect is typically soldered, RF connected, pogo pinned, zif connected or facilitated by some other mechanical device or means (not shown). Next, in FIG. 16b, there is shown a conventional arrangement in which the connection of a dipole 1606 is via a feed line 1605 to a balun 1604. The balun is disposed on PCB 1630′. The gap exists at the balun-to-feed system transition. This gap can be coupled to the RF system by micro-strip or strip-line across the gap. The balun 1604 is made large enough to establish an adequate amount of inductive reactance so that the system does not become too low in impedance.
Dual band gap split duplexer and/or matching is also a consideration. Balanced RF feed systems are often a consequence of symmetrical RF modules such as antennas, mixers, differential/push pull amplifiers, coplanar waveguides and other such devices. Solutions as described herein are applicable to all these areas, even thought the principle focus is for antenna applications and balun structures including inductive/transformer coupling.
With reference to FIGS. 17a-17e, conventional gap port feed systems are described. In FIG. 17a, it is gap 1702 across a balun 1701; in FIG. 17b, it is in a notch 1704 of a notch antenna 1703; and in FIG. 17c it is across the central region of a closed end slot 1706 of a notch antenna 1705.
The gap port defines the excitation region of the selected balanced RF system. The gap port is the subject of the transition from the balanced to the unbalanced RF circuit that needs to be connected to the antenna/balun. It should be understood that the edge opposite to the gap on the balun may be connected to a large or larger ground plane on which the RF circuits reside and still maintain the balanced/symmetrical condition. The balance remains as long as the attached ground plane is attached symmetrically to the balun even if it connects to the two adjacent sides as well.
FIG. 17d is an isometric view showing how a slot (or gap) 1708 in a ground plane is typically connected into a strip line 1709 feed system in an antenna system 1707. In the system 1707, the strip line 1709 connects to the opposite side of the gap/slot from where the line came from. In the side view of FIG. 17e, it is seen where the slot 1708 is coupled in a short circuit to the strip line 1709 at a point 1713 opposite on the slot. A similar configuration is shown in FIG. 17f, but in this case the strip line 1709′ passes across the slot 1708′ and beyond it by a distance of one quarter of a wavelength, ending in an open circuit. This well-practiced principle in strip line RF design shown in FIG. 17f effectively achieves a short circuit as in the location 1713 in FIG. 17e without the necessity of a direct electrical connection. This methodology finds its greatest use for the excitation of slot or notch antennas and also for the excitation of patch on slot antennas.